Javis OS
Javis OS turns the AI models you already pay for into a team of coworkers that share one memory, so switching to a better model never means starting over. Fourteen brains plug into the same toolbox, from Claude Code and Codex to OpenRouter, Gemini, Groq, DeepSeek, and local Ollama. Everything they learn lands in a Second Brain, an Obsidian-compatible markdown vault of memories, wiki pages, and sources, drawn as a knowledge graph you can search and edit. After each conversation, self-learning distills new memories, wiki entries, and skills, and every pass is a git commit you can undo in one tap or back up to a private GitHub repository. Hand over a goal in plain words and the Kanban dispatcher writes the spec, picks a worker, runs it in the background, and calls you only on exceptions. Loops and reminders fire on an interval, a clock time, or a cron expression. Coworkers on the Partners page get their own role, model, prompt, and skills, and teams of them chat in one room while a leader hands out work. Chatbots answer customers on WhatsApp, Messenger, Instagram, Telegram, Slack, and Zalo from one shared inbox. An MCP connection store holds several accounts per service under three hard-enforced permission levels. Hands-free voice, cost tracking, TOTP two-factor login, and in-app updates with rollback round it out. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.
Deploy